This is a study of residential patterns in urban America in 1940, taking advantage of newly available census data to pursue spatial questions at a level of detail and with methodologies that were not previously possible. It will organize the complete microdata file of 132 million Americans from the 1940 census into summary files at the level of the Enumeration District (ED), making it possible to study in detail the composition of small neighborhood areas across the country. These files will be made available at various geographic levels for the use of other researchers. For 69 major cities the EDs and census blocks will be mapped and households geocoded to ED segments of 1-3 blocks along a single street. For 6 metropolitan regions the individual records will be geocoded for both cities and suburbs, allowing more detailed analysis of residential patterns. These data will be used to study segregation patterns and processes that divided residents by race, ethnicity, and social class. 1) It will analyze variations across cities in the spatial assimilation of blacks (well intothe Great Migration from South to North) and immigrant groups in the first and second generation, plus segregation between families at different levels of SES. 2) It will examine the predictors of locational attainment to assess how race, ethnicity, class, and other characteristics affect the kinds of neighborhoods people live in, introducing characteristics of cities as contextual predictors of residential outcomes. 3) The geocoded address information will be exploited to study how individual residents are aggregated into neighborhoods in more detail, experimenting with a range of new methods to identify neighborhoods inductively based on their race/ethnicity and social class composition. 4) Geocoded data will also be used to draw overlapping, egocentric neighborhood areas around households at various spatial scales to define segregation patterns more flexibly. Discrete choice models will be applied to test hypotheses about the relative importance of race/ethnicity, class, and other attributes in constituting socialy distinct neighborhoods. These interrelated analyses will provide an important new point of reference for research on the incorporation of minority groups and immigrants in U.S. society today, as well as for the boundaries created by class segregation. The new spatial database will make a permanent and substantial addition to the nation's statistical infrastructure for health and population research, education, and policy making. It will allow study of neighborhood conditions for immigrants, minorities, and the urban poor at a critical historical moment and enable new kinds of spatial analysis. Linked to later demographic records and surveys it will allow study of neighborhood effects on life outcomes such as health, marriage and fertility, and longevity for the pre-baby boom generation of persons born before 1940.